Computational 3D imaging to quantify structural components and assembly of protein networks

A NanoFE Simulation-based Surrogate Machine Learning Model to Predict Mechanical Functionality of Protein Networks from Live Confocal Imaging

“…RNA isolation, molecular cloning and moss transfection were described previously in detail [24,25] and are therefore given here only in a shortened version. Total RNA was isolated from wild type Physcomitrella patens protonema using TRIzol Reagent (Thermo Fisher Scientific) and used for cDNA synthesis using Superscript III reverse transcriptase (Life Technologies, Carlsbad, CA, USA).…”

“…In 4 − 7 days after transfection, the protoplasts were concentrated to a volume of 100 µl, and 20 µl of this protoplast suspension was used for imaging. Confocal microscopy images (n = 37, i.e., 21 FtsZ2-1 and 16 FtsZ1-2 isoforms) were taken with a Leica TCS SP8 microscope (Leica Microsystems, Wetzlar, Germany), using the HCX PL APO 100x/1.40 oil objective and applying the microscopy conditions described previously [24,25]. A selection of images visualising FtsZ networks is depicted in Fig.…”

“…A set of 26 structural features describing the assembly of protein networks from global and local perspectives is extracted from each network. Here, only a short description of the workflow steps and features are given, details as well as a validation have been reported previously [24].…”

“…State-of-the-art microscopy imaging techniques permit resolving micro-structural details of protein networks. Computational analysis of acquired images facilitates the quantification of components and its assembly to networks [24], and may allow tracking structural changes of the network assembly triggered by internal or external stimuli, i.e., connecting the structure to functionality or distinguishing between network types [25]. Machine learning (ML) algorithms have proven to be remarkably capable for automating such complex image analysis tasks [26] and of correlating image content to biological structural functionality [27][28][29].…”

Feature‐based Classification of Protein Networks using Confocal Microscopy Imaging and Machine Learning

“…State-of-the-art imaging techniques allow resolving micro-structural details of protein networks. Computational analysis of these images permits resolving and quantifying the components and assembly of these networks [1], and may allow tracking structural changes in the network caused by internal or external stimuli, connecting the structure to functionality of the network or distinguishing between network types. Here, we present an automated method to classify protein networks based on their structural features exploiting a random forest model.…”